Predicting Subjective Impressions of the Roll Dynamics of Passenger Cars Using Characteristic Values in the Time and Frequency Domains

Subjective evaluations of a passenger car’s dynamics are often translated into quantitative, objective measures during the development of the vehicle, using a combination of characteristic values (CVs) and mathematical models. However, limited research exists for quantifying the subjective impression of the roll dynamics of passenger cars. This article presents a continuation of an initial study on the derivation of CVs to describe the roll dynamics of a BMW X5. In that work, correlation and regression analyses were used to determine CVs in the frequency domain for the following seven subjective criteria: roll stabilization at lower and higher frequencies, time delays at lower and higher frequencies, initial roll motion, roll angle overshoot, and overall impression. This article extends the initial study both by considering vehicle models from different market segments in the analysis of the original CVs and through deriving new CVs for the roll dynamics in the time domain. The resulting regression models were found to predict the subjective impression of a vehicle’s roll dynamics more accurately than those from the previous study, reducing the maximum absolute error across all models by 62% and the median error by up to 44%.